Rebuilding head map

[xsoliman]

Just reading the recent Seagate ST3500820AS 500GB Spindle seized thread
and wonder if anyone would care to enlighten me about head maps and making/re-building them

"Heads and platters swapped to good chassis in cleanroom environment, and after some clicking (normal) was able to make head map (H2 was a little weak) and image the folders required."
[presumably original pcb moved to the new chassis as well?]

"Head map on 7200.11 takes no time at all on UDMA "

"Through com it would, take a long time. But .11 head maps are done through ATA"


So was exactly is the head map in this context ?
Presumably, even with the original heads and platters (and pcb/rom), the fitting of them to a new chassis will not have exactly the same aligment as in the orginal chassis, and need to twek something (the head map and/or adaptives) to maximise the readability of the data

How is it made/rebuilt ?

Presumably PC3K has a function to do this, but how (broadly) does it do it eg contanstant reading of sectors on each zone and adjusting adaptives ?

What would be the sort of Seagate serial command to do it through the terminal ?

Thanks

[drc]

No, head map in this context is only referring to determining which LBAs exist on which physical heads

[xsoliman]

Thanks
But I'm still not sure why the head map would have to rebuilt/made in this pcb,head and platter swap case)


For those who dont know, the logical LBA's (of the user data section)
do not map directly to physical CHS on the disk, but are in various zones/maps
Which could potentially be different for each particular drive (if they are created at the manufacture / self scan stage ?) Or would they be the same for exact identical models ?

[pcimage]

The head map building didn't "have" to made in this case , but the building of the head map was purely precautionary, as I nearly always do in these type of cases where physical impact is involved, so as to identify a possble weak/failing head sooner rather than later.

If a weak head is identified (in this case nothing to do with alignment, but probably weakened in the impact) then I image the drive excluding the sectors pertaining to the suspect head as quickly as possible (i.e. image foward at UDMA speed), then image the remaining sectors at a slower speed (often backwards).